Retrieval of Trace Gas Concentrations from Lunar Occultation Measurements with Sciamachy on Envisat

L.K. Amekudzi, K. Bramstedt, A. Rozanov, H. Bovensmann, and J.P. Burrows

Abstract SCIAMACHY (Scanning Imaging Absorption Spectrometer for Atmospheric Chartography) is a UV-Vis-NIR spectrometer on-board ENVISAT, dedicated to measure atmospheric trace gas distributions in nadir, limb, and occultation geometries. In occultation mode, SCIAMACHY tracks the rising Sun or Moon directly through the atmosphere. Dividing the atmospheric measurements by extraterrestrial references gives transmission spectra, which are suitable for a DOAS-like retrieval approach. Here, we report on the major retrieved data products (ozone, NO2, and NO3) obtained from SCIAMACHY lunar occultation observations. We validated our ozone and NO2 data products. Ozone validation was carried out with other satellite instruments (HALOE, SAGE II, and POAM III). The NO2 results were validated with photochemically scaled HALOE and SAGE II profiles. The validation results reveal that the biases in the SCIAMACHY lunar occultation ozone and NO2 are within ±15%, with standard deviations and the uncertainties of the biases are less than 20% and 6%, respectively. Comparisons of NO3 results carried out with a 1-D photochemical model show that the accuracy of our NO3 data product is better than 30%.

1 Introduction

The SCanning Imaging Absorption SpectroMeter for Atmospheric CHartographY (SCIAMACHY) is a passive remote sensing moderate-resolution UV-Vis-NIR spectrometer on-board the European Space Agency's ENVIronmental SATellite (ENVISAT) launched into a near-circular Sun-synchronous polar orbit in March 2002 from Kourou, French Guiana. The instrument observes the Earth's atmosphere in nadir, limb, and solar/lunar occultation geometries, recording spectroscopic data in eight channels; covering the spectral range of 240-2380 nm with spectral

Institute of Environmental Physics and Remote Sensing (IUP/IFE), University of Bremen, Germany and Department of Physics, Kwame Nkrumah University of Science and Technology (KNUST), Kumasi, Ghana e-mail: [email protected]

resolution of 0.24-1.5 nm. Various atmospheric components (column and profile information of trace gases) are derived from SCIAMACHY measurements relevant to ozone chemistry, air pollution, and climate monitoring issues (Bovensmann et al. 1999; Gottwald et al. 2006).

Due to the Sun-synchronous orbit and the position of the SCIAMACHY instrument on ENVISAT, lunar occultation events are observed by SCIAMACHY in the Southern Hemisphere between 40°S and 90°S, when the phase of the Moon is 0.6-0.7 and ends shortly after full Moon, yielding measurements about 6-8 days per month. In the Northern Hemisphere between 49°N and 70°N, SCIAMACHY performs solar occultation measurements (Meyer et al. 2005; Bramstedt et al. 2007).

The SCIAMACHY lunar occultation measurements are successfully executed when the Moon's visibility is on the nightside of ENVISAT orbit, because on the day-side, stronger signals (stray light) from the brighter Earth's atmosphere contaminate the measurements. The duration of the useful SCIAMACHY lunar occultation measurements is highly variable (about 4-8 months per year), as seen in Fig. 1. The SCIAMACHY lunar occultation measurements are performed in Moon pointing mode, usually starting around 17 km the Moon is followed up to 100 km tangent height. The integration time for the lunar occultation measurements is 1.0 s resulting in a vertical sampling of approximately 2.5 km. The horizontal resolution is 30 km across track and extending approximately 400 km along track. Detailed information about SCIAMACHY lunar occultation is given in Amekudzi (2005) and Amekudzi et al. (2005).

The scientific objective of SCIAMACHY lunar occultation measurements is to provide nighttime vertical profiles of trace gases such as O3, NO2, and NO3, which are important in stratospheric chemistry. Simultaneous measurements of O3, NO2, and NO3 are crucial for understanding of NO* budget and long-term trends in stratospheric ozone loss.

Jan/03 Jul/03 Jan/04 Jul/04 Jan/05 Jul/05 Jan/06 Jul/06 Jan/07 Jul/07

Date

Fig. 1 Latitudinal distribution of SCIAMACHY lunar occultation measurements from 2003 to March 2007. Most measurements were taken in March, April, and May

Jan/03 Jul/03 Jan/04 Jul/04 Jan/05 Jul/05 Jan/06 Jul/06 Jan/07 Jul/07

Date

Fig. 1 Latitudinal distribution of SCIAMACHY lunar occultation measurements from 2003 to March 2007. Most measurements were taken in March, April, and May

In this paper, we describe briefly the retrieval method used to process the measured SCIAMACHY lunar occultation data. This is followed by O3, NO2, and NO3 retrieval results. Thereafter the validation carried out for the retrieved O3 and NO2 is presented, and finally, summary and conclusions of our findings.

2 Retrieval Methodology

In this section we described briefly the SCIAMACHY lunar occultation retrieval scheme. Detailed description of the retrieval method is given in Amekudzi (2005).

SCIATRAN version 2.1 radiative transfer and retrieval code (Rozanov et al. 2005) is used to process the calibrated (level-1, version 6.01) SCIAMACHY lunar occultation data to derive vertical profiles of O3, NO2, and NO3. The radiative transfer for simulating SCIAMACHY lunar transmission spectra and computing the Jacobian (the weighting function matrices) is based on Lambert-Beer's law. The simulated transmission Ys(hi, X) for a given tangent height, hi, and wavelength, X, is given by where AX is the total width of SCIAMACHY slit function S(X, X'), £ is the field of view of the instrument, and F(m) is the apparatus function. t(hi, X) is the full optical depth along the line of sight through the atmosphere and i represents the tangent height index.

The global fitting method coupled with the differential optical depth approach is applied to fit simultaneously NO2 and O3 within the spectral window of 420-454 nm and 520-580 nm, respectively. NO3 is fitted using the spectral window of 610680 nm containing the most intense NO3 absorption band at 623 nm V1(1,0) and 662 nm Vi(0,0). The NO3 retrieval window has significant contributions from other absorbers such as O3, O2, O4, and H2O. Therefore, to accurately fit and retrieve NO3 profiles, these interfering gases are fitted. O2 and H2O are line absorber, hence their absolute cross sections are calculated by using the exponential sum fitting of transmittance (ESFT) method (Buchwitz et al. 2000). Broadband absorption features of the atmosphere and instrument are removed from the measured spectra by subtracting a third order polynomial.

Typical spectral fits for NO2 and O3 are shown in Fig. 2 (top). Also in Fig. 2 (bottom) are shown the spectra residuals, which are the differences between the model differential spectra and the measurement contribution. The spectral residuals shown here are less than 0.5%.

The total model errors for the retrieval are within 2% and 10% for ozone and NO2 (Amekudzi 2005) and less than 20% for NO3 retrieval (Amekudzi et al. 2005). The total model errors include the uncertainties in the absorption cross sections for all interfering gases, which are fitted or retrieved with the targeted gases and errors due to tangent height shift, instrumental line shape, and temperature profiles.

wavelength [nm] wavelength [nm]

Fig. 2 Example of spectral fits (top) and residuals (bottom) at about 25 km tangent height showing absorption features of NO2 (top left) and O3 (top right) for measurement in orbit 15 598, day February 22, 2005. The diamond points represent the modeled differential optical depth and the solid line the measurement contribution wavelength [nm] wavelength [nm]

Fig. 2 Example of spectral fits (top) and residuals (bottom) at about 25 km tangent height showing absorption features of NO2 (top left) and O3 (top right) for measurement in orbit 15 598, day February 22, 2005. The diamond points represent the modeled differential optical depth and the solid line the measurement contribution

The European Centre for Medium-Range Weather Forecasts (ECMWF) temperature and pressure profiles information as well as O3 and NO2 absorption cross sections at five different temperatures measured at University of Bremen (Burrows et al. 1998) are used in the retrieval schemes. NO3 absorption cross section at 298 K (Sander et al. 2003) is used. The a priori trace gases are taken from model calculations from the Max-Planck Institute, Mainz. The inversion technique used in the retrieval algorithm is similar to the optimal estimation method (OEM) with additional smoothing constrain. Although the measurement vertical resolution is ~3-4 km, the retrievals were carried out with 1 km vertical sampling. The total retrieval errors, (i.e., the random, smoothing, and systematic errors) are within 5% and 15% between 18 km and 40 km for ozone (Amekudzi 2005). The retrieval errors for NO2 are within 5% and 20% in the altitude range of 18-36 km and for NO3 profiles the total retrieval errors are within 20% and 35% between 20 km and 45 km.

3 Retrieval Results

In this section we present zonal mean concentrations inferred from the retrieved number density profiles reported in Amekudzi et al. (2007b). In addition, monthly means of NO2 and NO3 are presented.

The zonal mean number densities derived from the retrieved lunar occultation O3, NO2, and NO3 vertical profiles are shown in Figs. 3, 4, and 5 respectively for 2003 (left) and 2004 (right). These results represent number density averages

Zonal Mean Ozone for 2003

Zonal Mean Ozone for 2004

45 40

45 40

I 35

Fig. 3 Zonal mean of ozone for 2003 and 2004 derived from SCIAMACHY lunar occultation measurements averaged over 1° latitude. Left are the 2003 results and right are 2004 results

45 40

-75 -70 -65 Latitude [degree]

Zonal Mean NO2 for 2003

Zonal Mean NO2 for 2003

10cm

10cm

Zonal Mean NO3 for 2003

Zonal Mean NO3 for 2004

50 45

25 20

corresponding to the latitudinal distributions shown in Fig. 1. There are 2-10 profiles per latitude bin between 85-76°S and between 75-60°S, there are 9-58 profiles per latitude bin. The 2003 results are derived from measurements for March to June and 2004 results for January to June.

In general higher values of ozone in the range of 3.0x1012 molec cm-3 to 4.5x 1012 molec cm-3 are observed for the altitude range 15-22 km. In 2004, much higher concentrations of ozone (4.5x1012 to 5.2x1012 molec cm-3) are retrieved for 70-84°S. Above 35 km very low values (< 0.5 x 1012 molec cm-3) of ozone are retrieved and between 22 km and 28 km the retrieved ozone concentrations are within 1.0 x 1012 molec cm-3 and 2.5 x 1012 molec cm-3.

The NO2 and NO3 results displayed in Figs. 4 and 5 show higher concentration values between latitudes 60° S and 65° S. The higher values of NO3 seen at these latitudes are mainly from measurements taken in March for 2003 and January to March for 2004 (see Fig. 7). March and April measurements contribute significantly to higher values of NO2 as seen at latitudes 60-65° S for 2003 and for NO2 results in 2004, higher values are from measurements in February to April (see Fig. 6). The maximum concentration of retrieved NO2 are in the range of 2.0 x 109 molec cm-3 to 4.0 x 109 molec cm-3 at altitudes of 25-35 km and those of NO3 are in the range of 1.2 x 107 molec cm-3 to 2.4 x 107 molec cm-3 at 34-42 km altitude.

NO Monthly Mean Profilas

NO Monthly Mean Profilas

0123458123458123456

NOs number density[x lo'cm 3]

Fig. 6 Monthly mean of NO2 for 2003 and 2004 derived from SCIAMACHY lunar occultation measurements. 2003 results in gray and 2004 in black. The error bars are the standard deviations of mean profiles

0123458123458123456

NOs number density[x lo'cm 3]

Fig. 6 Monthly mean of NO2 for 2003 and 2004 derived from SCIAMACHY lunar occultation measurements. 2003 results in gray and 2004 in black. The error bars are the standard deviations of mean profiles

NO Monthly Mean Profiles

NO Monthly Mean Profiles

0,0 0.5 1.0 1.5 2,0 2.5 3.CD.0 0,5 1.0 1.5 2.0 2.5 3.3) .0 0.5 1.0 1.5 2.0 2,5 3.0

N0s number density!* 10Tcm*

0,0 0.5 1.0 1.5 2,0 2.5 3.CD.0 0,5 1.0 1.5 2.0 2.5 3.3) .0 0.5 1.0 1.5 2.0 2,5 3.0

N0s number density!* 10Tcm*

Immediately after sunset, the stratospheric NO is rapidly oxidized to NO2 by O3. The NO2 formed, reacts more slowly with O3 to form NO3. The concentrations of both NO2 and NO3 thus build up within few hours after sunset. The NO2 and NO3 formed could react in the presence of a collision partner to form N2O5. In relatively warm stratosphere, N2O5 can be converted in the presence of collision partner to NO2 and NO3. At relatively cold temperatures N2O5 serves as the reservoir of the nighttime NOx or in the presence of polar stratospheric clouds N2O5 is converted to nitric acid (HNO3) and eventually to the stable complex nitric acid trihydrate (HNO3 ■ 3H2O). Kumer et al. (1997) have shown that at relative warm temperatures (T > 255 K), NO3 has lifetime of the order of a day and thermal lifetime of N2O5 is few minutes. This implies that more NO3 is produced in warmer nights. Hence our observations in January-March shown in Fig. 7 are consistent with nighttime NOy chemistry.

4 Validation Results

Retrieved profiles from remote sensing measurements require validation in order to access the overall confidence in the data products. Validation will help to detect and remove potential biases in the new retrieval products and also will provide information about estimated retrieval accuracy (Rodgers and Connor 2003). The major NO2 and O3 validation results are reported in Amekudzi et al. (2007a,c). Here we present a summary of the validation results. The reference validation data sources are retrieval results from the Halogen Occultation Experiment (HALOE), Stratospheric Aerosol and GAS Experiment II (SAGE II), and The Polar Ozone and Aerosol Measurement III (POAM III). The O3 data quality of these instruments is described in Cunnold et al. (1989); Brühl et al. (1996); Randall et al. (2003) and NO2 data quality is assessed in Cunnold et al. (1991); Gordley et al. (1996). The coincidences with HALOE are found in 2003-2005, SAGE II in 2004, and POAM in 2003-2004.

The statistics of O3 validation results (i.e., the mean relative deviations (mrd), the standard deviations of the mrd, and uncertainty) are shown in Table 1. The statistics of NO2 validation results are presented in Table 2. Due to the strong diurnal cycle of NO2, a photochemical correction scheme described in Bracher et al. (2005) has been applied to scale HALOE or SAGE II measurements to SCIAMACHY solar zenith angle. In general, very good agreements are obtained.

We compared our retrieved NO3 profiles with photochemical model calculations to check the validity and internal consistency of our results. Two photochemical model schemes were used. The first model scheme, called a full photochemical model is a 1-d chemistry model containing a comprehensive chemistry and dynamics of the stratosphere. Details of this model are described in Amekudzi et al. (2005) and references therein. The second model is a relatively simple chemical model, which assumed that at steady state, the nighttime concentration of NO3 depends on the concentration of ozone, NO2, temperature, and pressure. We found that NO3 profiles calculated from the full 1-D photochemical model are in good agreement with the retrieved NO3 profiles within 20% and 35% between 24 km and 45 km (Amekudzi 2005; Amekudzi et al. 2005). A comparison with a simple chemical

Table 1 Summary of lunar occultation ozone validation results, the coincidence criteria applied are measurement time difference of 12 h and correlative radius of 1000 km

Instruments

mrd (%)

rms (%)

Uncert.a (%)

Height (km)

HALOE N = 154

-5 to +15

5-25

2-5

20-45

SAGE IIN = 92

-15 to+15

6-20

< 2

20-45

POAM III N = 149

-8 to+2

12-20

1-4

24-43

a The uncertainty is the rms/*JN, where N is the number of coincidences.

a The uncertainty is the rms/*JN, where N is the number of coincidences.

Table 2 Summary of lunar occultation NO2 validation results. The coincidence criteria applied are similar to the ozone validation

Instruments

mrd (%)

rms (%)

Uncert. (%)

Height (km)

HALOE N =

65

-16 to +3

4-16

1-6

25-38

SAGE II N =

72

-9 to +7

10-17

<2

22-39

model calculation showed good agreement between 22 km and 38 km with accuracy better than 30% (Amekudzi et al. 2005, 2007a).

5 Summary and Conclusions

The retrieval results of SCIAMACHY lunar occultation measurements of O3, NO2, and NO3 are presented. Relatively high concentration values were retrieved for NO3 in the months of January-March due to warmer temperatures. These results are consistent with nighttime NOy chemistry.

Validation performed with correlative satellite instruments showed that the SCIAMACHY lunar occultation O3 and NO2 results are very promising. The biases in the O3 and NO2 validation were within ±15%, with standard deviations and the uncertainties of the biases better than 20% and 6%, respectively. Comparisons of NO3 results carried out with a 1-D photochemical model showed that the accuracy of our NO3 data products were better than 30%. Re-processing of the complete dataset and validation activities are currently in progress. Comparisons of our NO3 profiles with stellar occultation NO3 measurements from GOMOS (Global Ozone Monitoring by Occultation of Stars) instrument will be considered in the future.

Acknowledgements We are thankful to the following institutions: European Space Agency (ESA) for providing SCIAMACHY level-1 data. The HALOE and SAGE II teams at Hampton University and NASA Langley Research Center (LaRC), USA, for providing us with HALOE and SAGE II data. The ECMWF special project SPDECDIO provided the temperature and pressure profiles for this work. This work was funded in parts by the German Ministry of Education and Research (BMBF) via grant 07UFE12/8, the DLR-Bonn via grant 50EE0502, the University of Bremen and the state of Bremen. We are grateful to our anonymous reviewers for their helpful comments and suggestions.

References

Amekudzi LK (2005) Stratospheric O3, NO2, and NO3 number density profiles from SCIAMACHY lunar occultation spectroscopic measurements: Retrieval, validation and interpretation. PhD thesis, Universitat Bremen, ISBN 3-8325-1131-8, Logos Verlag Berlin Amekudzi LK, Sinnhuber BM, Sheode NV, Meyer J, Rozanov A, Lamsal LN, Bovensmann H, Burrows JP (2005) Retrieval of stratospheric NO3 vertical profiles from SCIAMACHY lunar occultation measurement over the Antarctic. J Geophys Res 110(D20304), doi:10.1029/2004JD005748 Amekudzi LK, Bracher A, Bramstedt K, Rozanov A, Bovensmann H, Burrows JP (2007a) Towards validation of SCIAMACHY lunar occultation NO2 vertical profiles. Adv Space Res 41(11):1921-1932, doi:10.1016/j.asr.2007.06.055 Amekudzi LK, Bramstedt K, Bracher A, Rozanov A, Bovensmann H, Burrows JP (2007b) Current retrieval and intercomparisons results of SCIAMACHY nighttime NO^. In: Proc. of Envisat Symposium 2007, ESA Publication Division, ESTEC, Noordwijk, The Netherlands, SP-636 Amekudzi LK, Bramstedt K, Bracher A, Rozanov A, Bovensmann H, Burrows JP (2007c) SCIA-MACHY solar and lunar occultation: validation of ozone, NO2 and NO3 profiles. In: Proc of Atmospheric Chemistry Validation (ACVE-3) December 2006, ESA Publication Division, ESTEC, Noordwijk, The Netherlands, SP-642

Bovensmann H, Burrows JP, Buchwitz M, Frerick J, Noel S, Rozanov VV, Chance KV, Goede APH (1999) SCIAMACHY: Mission objectives and measurement modes. J Atmos Sci 56(2): 127-150, doi:10.1175/1520-0469(1999)056¡0127:SMOAMM¿2.0.CO•,2 Bracher A, Sinnhuber M, Rozanov A, Burrows JP (2005) Using a photochemical model for the validation of NO2 satellite measurements at different solar zenith angles. J Atmos Chem Phys 5:393-408

Bramstedt K, Amekudzi LK, Bracher A, Rozanov A, Bovensmann H, Burrows JP (2007) SCIA-MACHY solar occultation: Ozone and NO2 profiles 2002-2006. In: Proc. of Envisat Symposium 2007, ESA Publication Division, ESTEC, Noordwijk, The Netherlands, SP-636 Brühl C, Drayson SR, Russell JM, Crutzen PJ, Mclnerney JM, Purcell PN, Claude H, Gernandt H, McGee TJ, McDermid IS, Gunson MR (1996) Halogen occultation experiment ozone channel validation. J Geophys Res 101(D6):10 217-10 240, doi:10.1029/95JD02031 Buchwitz M, Rozanov VV, Burrows JP (2000) A correlated-^ distribution scheme for overlapping gases suitable for retrieval of atmospheric constitunents from moderate resolution radiance measurements in the visible/near-infrared spectral region. J Geophys Res 105(D12): 15 247-15 262, doi:10.1029/2000JD900171 Burrows JP, Dehn A, Deters B, Himmelmann S, Richter A, Voigt S, Orphal J (1998) Atmospheric remote-sensing reference data from GOME: Part 1. Temperature-dependent absorption cross sections of NO2 in the 231-794 nm range. J Quant Spectrosc Radiat Transf 60(6):1025-1031 Cunnold DM, Chu WP, Barnes RA, McCormick MP, Veiga RE (1989) Validation of SAGE II

ozone measurements. J Geophys Res 94:8447-8460 Cunnold DM, Zawodny JM, Chu WP, Pommereau JP, Goutail F, Lenoble J, McCormick MP, Veiga RE, Murcray D, Iwagami N, Shibasaki K, Simon PC, Peetermans W (1991) Validation of SAGE II NO2 measurements. J Geophys Res 96(D7):12913-12925 Gordley LL, Russell JMI, Mickley LJ, Frederick JE, Park JH, Stone KA, Beaver GM, McInerney JM, Deaver LE, Toon GC, Murcray FJ, Blatherwick RD, Gunson MR, Abbatt JPD, Mauldin RLI, Mount GH, Sen B, Blavier JF (1996) Validation of nitric oxide and nitrogen dioxide measurements made by the halogen occultation experiment for the UARS platform. J Geophys Res 101(D6):10241-10266 Gottwald M, Bovensmann H, et al. (2006) SCIAMACHY monitoring the changing Earth's atmosphere. DLR, Germany, Institut fur Methodik der Fernerkundung (IMF) Kumer JB, Kawa SR, Roche AE, Mergenthaler JL, Smith SE, Taylor FW, Connell PS, Douglass AR (1997) UARS first global N2O5 data set: Application to a stratospheric warming event in January 1992. J Geophys Res 102(D3):3575-3582 Meyer J, Bracher A, Rozanov A, Schlesier AC, Bovensmann H, Burrows JP (2005) Solar occultation with SCIAMACHY: algorithm description and first validation. J Atmos Chem Phys 5:1589-1604

Randall CE, Rusch DW, Bevilacqua RM, Hoppel KW, Lumpe JD, Shettle E, Thompson E, Deaver L, Zawodny J, Kyro E, Johnson B, Kelder H, Dorokhov VM, Konig-Langlo G, Gil M (2003) Validation of POAM III ozone: Comparisons with ozonesonde and satellite data. J Geophys Res 108(D12):4367, doi:10.1029/2002JD002944 Rodgers CD, Connor BJ (2003) Intercomparison of remote sounding instruments. J Geophys Res

108:4116, doi:10.1029/2002JD002944 Rozanov A, Rozanov V, Buchwitz M, Kokhanovsky A, Burrows JP (2005) SCIATRAN 2.0 - A new radiative transfer model for geophysical applications in the 175-2400 nm spectral region. Adv Space Res 36(5):1015-1019 Sander SP, et al. (2003) Chemical kinetics and photochemical data for use in stratospheric modeling. NASA Panel Data Eval 14:1-119-4-111, JPL Publication 02-25

0 0

Post a comment